[ElevenLathe]

I've played with GT and even pull it out every once in a while for small tasks, though I haven't in about 6 months.

I think the idea is good, but it's a tough sell for working programmers because the whole culture of it is so foreign. I think there's a version of GT that would do well if it described itself in terms of paradigms working coders know (POSIX, IDEs, blub-y languages, text files) instead of in terms of SmallTalk. Maybe something nearly as cool can be done as a VSCode plugin? I personally think of it as kind of a supercharged Emacs for SmallTalkers.

[tudorgirba]

Thank you for giving it a try and suggestion.

The first goal was to help us explore how far can the idea of contextual tools go. It helped discover what today we call Moldable Development. It is also the first extensive case study of Moldable Development, itself offering 5+K contextual tools that we used to develop the environment itself. And when we work on a system, we build thousands more.

That said, now that we know what Moldable Development is, it can be copied. We want people to copy it. Our worry though is that we want people to copy everything, not only the visible parts.

For example, I understand the Emacs parallel. But think of this: while Emacs can be extended, how many extensions do you actually use that are specific to your system? We literally use thousands. Per system. That quantitative difference leads to a qualitative difference and it's made possible because of the totality of the environment.

[ElevenLathe]

I think a supercharger changes a passenger car engine qualitatively too!

So are there plans to copy GT/moldable development somewhere outside of Pharo/Smalltalk?

[tudorgirba]

Moldable Development is not specific to Smalltalk. We already show how it works for other languages. Some ideas are already inspiring other environments, like Clerk.

In the meantime, consider GT as an extensive blueprint of what's possible. If there is one thing we learnt is that the technology is the smallest investment. The real investment is in learning how to exploit the idea of contextual tools for solving hard problems. That's what takes the longest, but the difference to how those problems are approached today can be measured in orders of magnitude.